Prune wrote:

What's the best tube type to use for a voltage controlled current sink,
with respect to linearity?

I am building plasma tweeters, with a DC glow discharge modulated by audio
(similar to US patent 4,219,705, but I'm using microhollow discharges
(MHCD) as cathodes for the larger discharge to remove cathode fall
instabilities; this technique was unknown during the time of that patent).
My power supply is 2700 V at 400 mA (200 mA for each of two channels for a
stereo setup), ripple filtered to 0.01 V. I'm hoping to have the discharge
drop most of the voltage, and the driving device a lesser portion. The
driver needs to be a voltage controlled current sink at the plasma cathode
(the plasma anode is the high voltage). Since the audio modulation is only
a few percent of the bias, less than 20 mA, essentially I'm concerned with
linearity in a small range around the average operating point. I need to
know what's the best tube type to use in terms of the sound quality, and in
what configuration (grounded cathode or grid, etc.). Most current drives
use pentodes, but I've been told a triode is more linear.

I don't know the best tube for your application, but I _can_ tell you
that a triode has a quite significantly lower plate resistance than a
pentode; if you really need the current to be independent of plate
voltage then a triode without heavy cathode degeneration isn't the way
to go.

I believe that triodes are more linear when driving a resistive load
exactly because of this lower plate impedance -- if you hold the plate
voltage constant then the plate current vs. grid voltage relationship
isn't any better than a pentode, but if you allow the plate voltage to
vary it provides internal feedback on the tube current.

Hopefully you'll get someone who'll give you a more definitive answer.
Perhaps looking for a newsgroup that involves audio, rather than radio,
homebrewing will help.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com